Devirtualize numel() #27294
Closed
Devirtualize numel() #27294
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Fixes #27291 I think this deletes at::numel function too, making it technically BC-breaking, but I don't think we intended to generate that function in the first place. Signed-off-by: Edward Z. Yang <[email protected]> [ghstack-poisoned]
pytorchbot
added
caffe2
module: internals
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Oct 3, 2019
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> ghstack-source-id: 53d0e3e Pull Request resolved: #27294
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> [ghstack-poisoned]
Contributor
Author
|
@ailzhang this one breaks xla but it should be a pretty simple fix; just delete the numel() implementation. (This is under the assumption that your TensorImpl::numel implementation does the right thing. Which it should.) |
Contributor
|
Thanks @ezyang , I'm starting a local run to verify the change. :D Will report back once it's finished. |
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Oct 3, 2019
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> ghstack-source-id: 9ce95a1 Pull Request resolved: #27294
ailzhang
reviewed
Oct 3, 2019
| return [sorted(g, key=declkey) for g in grouped_decls] | ||
|
|
||
| # We need to add methods implemented manually in TensorImpl | ||
| # TODO: This seems to claim sizes() returns an int64_t. Really? |
Contributor
There was a problem hiding this comment.
our Tensor::sizes() does return a int64_t iirc. It's number of dimensions. And yes I agree it's confusing :P
ailzhang
approved these changes
Oct 3, 2019
Contributor
ailzhang
left a comment
ailzhang
left a comment
There was a problem hiding this comment.
Fix is in pytorch/xla#1136
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Oct 3, 2019
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> ghstack-source-id: e65c8ab Pull Request resolved: #27294
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> Differential Revision: [D17763486](https://our.internmc.facebook.com/intern/diff/D17763486) [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Oct 4, 2019
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> ghstack-source-id: 74091b8 Pull Request resolved: #27294
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> Differential Revision: [D17763486](https://our.internmc.facebook.com/intern/diff/D17763486) [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Oct 4, 2019
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> ghstack-source-id: 7b0973e Pull Request resolved: #27294
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> Differential Revision: [D17763486](https://our.internmc.facebook.com/intern/diff/D17763486) [ghstack-poisoned]
ezyang
added a commit
that referenced
this pull request
Oct 8, 2019
Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> ghstack-source-id: 71653a9 Pull Request resolved: #27294
Contributor
zdevito
pushed a commit
to zdevito/ATen
that referenced
this pull request
Oct 9, 2019
Summary: Pull Request resolved: pytorch/pytorch#27294 Fixes #27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> Test Plan: Imported from OSS Differential Revision: D17763486 Pulled By: ezyang fbshipit-source-id: 5793b53e2db80b044e57faae325a95c649d9d459
thiagocrepaldi
pushed a commit
to thiagocrepaldi/pytorch
that referenced
this pull request
Feb 4, 2020
Summary: Pull Request resolved: pytorch#27294 Fixes pytorch#27291 I'm a little annoyed that I have to reintroduce manual binding code. But it's probably not a good idea to teach the codegen how to do fastpath functions (is it?) Signed-off-by: Edward Z. Yang <[email protected]> Test Plan: Imported from OSS Differential Revision: D17763486 Pulled By: ezyang fbshipit-source-id: 5793b53e2db80b044e57faae325a95c649d9d459
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack:
Fixes #27291
I'm a little annoyed that I have to reintroduce manual binding code. But it's
probably not a good idea to teach the codegen how to do fastpath functions
(is it?)
Signed-off-by: Edward Z. Yang [email protected]
Differential Revision: D17763486